Flow control

Flow control statements allow to:

Flow control

Sequential execution

In the simplest scenario a program is executed line-by-line:

statement_1
statement_2
statement_3

Note, sometimes several statements might be in the same line separated with semicolon:

statement_1
statement_2; statement_3

Conditional execution

if

Runs statement_T only when the condition is TRUE.

statement_1

if( condition ) {
  # this runs only when the condition is TRUE
  statement_T
}

statement_2

if/else

When the condition is TRUE runs statement_T.
Otherwise runs statement_F.

statement_1

if( condition ) {
  # this runs only when the condition is TRUE
  statement_T
} else {
  # this runs only when the condition is FALSE
  statement_F
}

statement_2

if/else if/…/else

Multiple conditions can be combined:

statement_1

if ( condition1 ) {
  # this runs only when the condition1 is TRUE
  statement_T1
} else if( condition2 ) {
  # this runs only when the condition1 was FALSE and the condition2 is TRUE
  statement_T2
} else {
  # this runs only when all above conditions were FALSE
  statement_F
}

statement_2

Looping

for loop

Executes the same statement a fixed number of times:

for( v in 1:5 ) {
  # this runs once for v==1, then v==2, ..., v==5
  print( v )
}
[1] 1
[1] 2
[1] 3
[1] 4
[1] 5

Or allows to iterate over each element of a vector:

fruits <- c( "apples", "bananas", "cherries" )
for( fruit in fruits ) {
  print( fruit )
}
[1] "apples"
[1] "bananas"
[1] "cherries"

Or for each element of a list:

studentScores <- list(
  Amy = c( 4,8,6,9,7 ),
  Bob = c( 3,5,5,4 ),
  Chester = c( 7,7,6,6,9,5,6,8 )
)
meanScores <- c()
for( student in names( studentScores ) ) {
  meanScores[[ student ]] <- mean( studentScores[[ student ]] )
}
meanScores
    Amy     Bob Chester 
   6.80    4.25    6.75

Note, the above example would rather be implemented with sapply:

meanScores <- sapply( studentScores, mean )
meanScores
    Amy     Bob Chester 
   6.80    4.25    6.75

Example: a function with if/else statements

Let’s take the following from a World Health Organization document:

Now we would like to write a function ageGroup which takes age argument (in years) and returns the age group label.

First let’s enumerate the age group categories (for a country where 19 years is the adult age):

group condition for age [y]
adult age > 19
adolescent age >= 10 && age <= 19
child age >= 1 && age < 10
infant age < 1

Function template

Here is a template of the ageGroup function:

ageGroup <- function( age ) { 
  # age: age given in years
  ...
}

Here are expected results:

ageGroup(0)   # "infant"
ageGroup(10)  # "adolescent"
ageGroup(9)   # "child"
ageGroup(20)  # "adult"
ageGroup(-1)  # should generate an error

Poor implementation with if only

Let’s first implement a dummy function which always returns child; of course this would provide wrong results:

ageGroup <- function( age ) { 
  # age: age given in years
  result <- "child"
  result
}
ageGroup(0)   # "infant"
[1] "child"
ageGroup(10)  # "adolescent"
[1] "child"
ageGroup(9)   # "child"
[1] "child"
ageGroup(20)  # "adult"
[1] "child"
ageGroup(-1)  # should generate an error
[1] "child"

Next we add a single conditional to capture adult label:

ageGroup <- function( age ) { 
  # age: age given in years
  result <- "child"
  if( age > 19 ) {
    result <- "adult"
  }
  result
}
ageGroup(0)   # "infant"
[1] "child"
ageGroup(10)  # "adolescent"
[1] "child"
ageGroup(9)   # "child"
[1] "child"
ageGroup(20)  # "adult"
[1] "adult"
ageGroup(-1)  # should generate an error
[1] "child"

Finally we add the conditional statements to capture the labels adolescent and infant in the same way:

ageGroup <- function( age ) { 
  # age: age given in years
  result <- "child"
  if( age > 19 ) {
    result <- "adult"
  }
  if( age >= 10 & age <= 19 ) {
    result <- "adolescent"
  } 
  if( age < 1 ) {
    result <- "infant"
  } 
  result
}
ageGroup(0)   # "infant"
[1] "infant"
ageGroup(10)  # "adolescent"
[1] "adolescent"
ageGroup(9)   # "child"
[1] "child"
ageGroup(20)  # "adult"
[1] "adult"
ageGroup(-1)  # should generate an error
[1] "infant"

To generate the error we request the program to stop with an error message:

ageGroup <- function( age ) { 
  # age: age given in years
  result <- "child"
  if( age > 19 ) {
    result <- "adult"
  }
  if( age >= 10 & age <= 19 ) {
    result <- "adolescent"
  } 
  if( age < 1 ) {
    result <- "infant"
  } 
  if( age < 0 ) {
    stop( "Age can't be negative." )
  } 
  result
}
ageGroup(0)   # "infant"
[1] "infant"
ageGroup(10)  # "adolescent"
[1] "adolescent"
ageGroup(9)   # "child"
[1] "child"
ageGroup(20)  # "adult"
[1] "adult"
ageGroup(-1)  # should generate an error
Error in ageGroup(-1): Age can't be negative.

Better implementation with if/else

In the previous scenario: when age is 30 and the result is set to adult it makes no sense anymore to check whether age is in other ranges.
if/else construction allows for that.

ageGroup <- function( age ) { 
  # age: age given in years
  if( age > 19 ) {
    result <- "adult"
  } else if( age >= 10 ) {
    result <- "adolescent"
  } else if( age >= 1 ) {
    result <- "child"
  } else if( age >= 0 ) {
    result <- "infant"
  } else {
    stop( "Age can't be negative." )
  }
  result
}
ageGroup(0)   # "infant"
[1] "infant"
ageGroup(10)  # "adolescent"
[1] "adolescent"
ageGroup(9)   # "child"
[1] "child"
ageGroup(20)  # "adult"
[1] "adult"
ageGroup(-1)  # should generate an error
Error in ageGroup(-1): Age can't be negative.

Copyright © 2021 Biomedical Data Sciences (BDS) | LUMC